The work is devoted to the study of the importance of the primary level and dynamics of clinical, immunological and virological indicators of HIV status in order to further predict the probability of developing risks of an unfavorable course of the disease and the effectiveness of highly active antiretroviral therapy (HAART) for individualization and improvement of management tactics for HIV-infected patients. A comprehensive clinical and laboratory examination was performed and risk factors for developing an unfavorable prognosis of the disease were identified in 450 HIV-infected patients. The main reasons of losses of patients in the 5-year follow − up cohort have been identified, including discontinuation of HAART −15,3%, death from HIV − related causes – 14,0%, death from non-HIV-related causes and IRSR – 3,1%. Taking into account the results obtained, a correlation analysis of clinical and laboratory parameters of a cohort of HIV-infected patients was performed, according to the results of which 10 main risk factors were selected from 22 studied indicators. Thus, it was found that the unfavorable course of HIV infection is influenced by the following factors: age over 40 years (r =0,14, p <0,003), male gender (r =0,18, p <0,007), 4th clinical stage of HIV infection (r =0,26, p <0,000), late appointment of ART due to late revealing (r =0,16, p <0,000), high initial viral load of HIV RNA (r =0,22, p <0,001), HBV coinfection (r =0,12, p <0,022), low HAART compliance (r =0,48, p <0,000), low initial CD4+T cell levels (r =-0,30, p <0,000), and CD4+T cell levels after 6 (r =-0,29, p <0,000) and 12 (r =-0,20, p <0,005) months of HAART. To develop a predictive mathematical model of an unfavorable prognosis of HIV infection, data from the previous correlation analysis were used, which identified the main risk factors for an unfavorable course of the disease, Cox regression analysis and ROC analysis were used to determine cumulative risks. Using sequential Wald analysis, evaluation scores were determined for each of the 10 indicators. Thus, the presence of 4th clinical stage of HIV, male gender and age of the patient over 40 years were assessed at +2 points, the level of CD4+T-lymphocytes before the start of therapy < 160 CL./µL – Y +3 points, Lg HIV RNA before the start of therapy >4.88 copies/ml – y +2 points, CD4+T-lymphocyte levels after 6 and 12 months of therapy < 228 CL. / MCL – u +4 points unmotivated discontinuation of art – U +9 points. The constructed mathematical model makes it possible to predict a high risk of mortality based on a total score of 10 indicators with a predictive accuracy of 89,7%; sensitivity – 76,2%; specificity – 94,0%. For more convenient use in practice, an «algorithm for calculating the probability of an unfavorable increase in HIV infection in patients taking HAART» was created. Thus, the mathematical prognostic model created on the basis of initial socio-demographic and clinical-laboratory indicators allows for an early prognosis of the unfavorable course of HIV infection in patients at the time of HAART administration and with dynamic follow-up after 6 and 12 months of therapy. The model contains a small number of parameters, does not require complex calculations and has a fairly high indicator of prognostic effectiveness, which makes it possible to individualize the management tactics of patients with HIV infection.
